Social impairments are core features of schizophrenia that lead to poor outcome. Social skills and competence improve quality of life and protect against stress-related exacerbation of symptoms, while supporting resilience, interpersonal interactions, and social affiliation. To improve outcome, we must remediate social deficits. Existing psychosocial interventions are moderately effective but the effort-intensive nature (high burden), low adherence, and weak transfer of skills to everyday life present significant hurdles toward recovery. Thus, there is a dire need to develop effective, engaging and low-burden social interventions for people with schizophrenia that will result in better compliance rates and functional outcome. We will test the effectiveness of a novel adaptive virtual reality (VR) intervention in improving targeted social cognitive function (social attention as indexed by eye scanning patterns) in individuals with schizophrenia. VR technology offers a flexible alternative to conventional therapies, with several advantages, including a simplified and low-stress social interaction environment with targeted opportunities to simulate, exercise and reinforce basic elements of social skills in a very wide range of realistic scenarios, and to repea exposure to naturalistic situations from multiple angles. Our desktop VR `game' is designed as an `intelligent' system that adaptively adjusts the difficulty of social training tasks based on participant's physiological, eye tracking and performance data in real time. Such dynamic feedback-based, `closed-loop' VR supports and enhances training because it adjusts and personalizes the learning environment in real time for each participant so that he/she always learns at an optimal arousal and attentional state. Furthermore, the VR environment can potentially simulate any social scenarios, which allows the participants to exercise social skills n a wide variety of situations. Such simulation exercises can help generalize learned skills to everyday life. The R21 phase will aim to implement the social intervention VR task, test its efficacy on improving social attention (target) in schizophrenia, and determine an optimal `dose'. We hypothesize that the VR training will engage social attention and improving social attention will lead to better social outcome. The R33 phase will test the adaptive social VR game against an active control condition in a pilot randomized controlled trial to evaluate the relative efficac of the social VR on enhancing social attention and associated neural circuitry. We will also examine social outcome. If this initial work is successful, our long-term goals are to develop VR social skills training modules that are personalized, accessible, and portable so that social remediation can become an integral part of one's daily life.